Intelligent pausing of recording by a property monitoring system

1. A computer-implemented method comprising:

receiving, from a camera, a first video clip indicative of an event;

determining, from the first video clip, a category for the first video clip;

determining, for the first video clip and based on the category, a difference threshold;

receiving, from the camera, a second video clip;

determining, for the first video clip and the second video clip, a difference score; and

in response to determining that the difference score for the first video clip and the second video clip meets the difference threshold, providing an alert.

2. The method of claim 1 , wherein determining, from the first video clip, a category for the first video clip comprises:

determining an action being performed by a human in the first video clip; and

determining the category for the first video clip based on the action being performed by the human.

3. The method of claim 1 , wherein determining an action being performed by a human in the first video clip comprises:

identifying a particular object in the first video clip;

determining that the human is interacting with the particular object in the first video clip; and

determining the action being performed by the human in the first video clip based on the particular object with which the human was determined to be interacting.

4. The method of claim 1 , wherein determining, for the first video clip and the second video clip, a difference score comprises:

determining that a human present in the second video clip was not present in the first video clip; and

based on determining that the human present in the second video clip was not present in the first video clip, determining the difference score.

5. The method of claim 1 , wherein determining, for the first video clip and the second video clip, a difference score comprises:

determining, from the second video clip, a second category for the second video clip; and

determining the difference score based on a difference between the category for the first video clip and the second category for the second video clip.

6. The method of claim 1 , wherein determining, for the first video clip and based on the category, a difference threshold comprises:

determining a risk level based on the category; and

determining the difference threshold based on the risk level.

7. The method of claim 6 , wherein determining the difference threshold based on the risk level comprises:

determining, from the first video clip, a second category for the first video clip;

determining a second risk level based on the second category; and

determining the difference threshold based on the risk level and the second risk level.

8. The method of claim 1 , wherein the first video clip and the second video clip are both a predetermined duration.

9. A system comprising:

one or more computers and one or more storage devices storing instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:

maintaining, for each of a plurality of categories for content detected in respective video clips, a corresponding difference threshold, a plurality of difference thresholds including the corresponding difference thresholds at least some of which have different values than another difference threshold;

receiving, from a camera, a first video clip indicative of an event;

determining, from the plurality of categories and using the first video clip, a category for content depicted in the first video clip;

determining, for the first video clip and using the category, a difference threshold from the plurality of difference thresholds;

receiving, from the camera, a second video clip;

determining, for the first video clip and the second video clip, a difference score that represents a difference between at least a first portion of the first video clip and at least a second portion of the second video clip;

determining whether the difference score for the first video clip and the second video clip satisfies the difference threshold; and

in response to determining that the difference score for the first video clip and the second video clip satisfies the difference threshold, providing an alert.

10. The system of claim 9 , wherein determining, from the first video clip, a category for the first video clip comprises:

determining an action being performed by a human in the first video clip; and

determining the category for the first video clip based on the action being performed by the human.

11. The system of claim 9 , wherein determining an action being performed by a human in the first video clip comprises:

identifying a particular object in the first video clip;

determining that the human is interacting with the particular object in the first video clip; and

determining the action being performed by the human in the first video clip based on the particular object with which the human was determined to be interacting.

12. The system of claim 9 , wherein determining, for the first video clip and the second video clip, a difference score comprises:

determining that a human present in the second video clip was not present in the first video clip; and

based on determining that the human present in the second video clip was not present in the first video clip, determining the difference score.

13. The system of claim 9 , wherein determining, for the first video clip and the second video clip, a difference score comprises:

determining, from the second video clip, a second category for the second video clip; and

determining the difference score based on a difference between the category for the first video clip and the second category for the second video clip.

14. The system of claim 9 , wherein determining, for the first video clip and based on the category, a difference threshold comprises:

determining a risk level based on the category; and

determining the difference threshold based on the risk level.

15. The system of claim 14 , wherein determining the difference threshold based on the risk level comprises:

determining, from the first video clip, a second category for the first video clip;

determining a second risk level based on the second category; and

determining the difference threshold based on the risk level and the second risk level.

16. The system of claim 9 , wherein the first video clip and the second video clip are both a predetermined duration.

17. A non-transitory computer-readable medium storing software comprising instructions executable by one or more computers which, upon such execution, cause the one or more computers to perform operations comprising:

receiving, from a camera, a first video clip indicative of an event;

determining, from the first video clip, a category for the first video clip;

determining, for the first video clip and based on the category, a difference threshold, wherein determining the difference threshold comprises:

determining a risk level based on the category; and

determining the difference threshold based on the risk level;

receiving, from the camera, a second video clip;

determining, for the first video clip and the second video clip, a difference score, wherein determining the difference score comprises:

determining, from the second video clip, a second category for the second video clip;

determining that a human present in the second video clip was not present in the first video clip; and

based on (i) a difference between the category for the first video clip and the second category for the second video clip and (ii) determining that the human present in the second video clip was not present in the first video clip, determining the difference score; and

in response to determining that the difference score for the first video clip and the second video clip meets the difference threshold, providing an alert.

18. The medium of claim 17 , wherein determining, from the first video clip, a category for the first video clip comprises:

determining an action being performed by a human in the first video clip; and

determining the category for the first video clip based on the action being performed by the human.

19. The medium of claim 17 , wherein determining an action being performed by a human in the first video clip comprises:

identifying a particular object in the first video clip;

determining that the human is interacting with the particular object in the first video clip; and

determining the action being performed by the human in the first video clip based on the particular object with which the human was determined to be interacting.